Glaucoma Detection Based On Shape Features and SMOTE Algorithm

Abstract

Glaucoma is a chronic eye condition in which the optic nerve is progressively damaged. As the disease progresses, more optic head damage due to loss of peripheral vision and a resultant in gradually vision loss. Glaucoma progression precedes some structural damage in the retina are signs of glaucoma appear as changes in size, structure, shape, and color of the optic disc and optic cup, which suffer from the subjectivity of human due to experience, fatigue factor etc. There are increasing demands for medical image-based Computer-Aided Diagnosis (CAD) systems based on fundus image for glaucoma detection, thus the human mistakes, other retinal diseases like Age-related Macular Degeneration (AMD) affecting in early glaucoma detection, and the existing medical devices like Optical Coherence Tomography (OCT) and Heidelberg Retinal Tomography (HRT) are expensive. This paper proposes a novel algorithm by extract 13 shape features from disc and cup. Next, the best features selected using the student t-test method and balanced using SMOTE algorithm. The evaluation of the proposed algorithm is performed using a RIM_ONE database, the average accuracy 91.3%, maximize the area under the curve (AUC) 0.92, using Ensembles RUSBoosted tree. Future works suggested designing a complete, automated CAD system using a different type of features.